Unemployment and crime since 1920

The relationship between the labour market and criminal behaviour has been of longstanding interest in the social sciences. The intuitive appeal of the argument that improving labour market conditions causes individuals to commit less crime, while apparently self-evident, is supported by empirical evidence, mostly focusing on the post WWII period. Typical analyses have exploited geographical variation in crime rates and labour market opportunities to estimate the effect of unemployment, while controlling for a host of confounding variables, or using various combinations of instrumental variables to establish the direction of causality.
In the Netherlands, labour market prospects fluctuated strongly from the start to the conclusion of the twentieth century. During the Great Depression, after 1930, the unemployment rate was around 20 per cent, from which it declined to around 3 per cent for most of the 1950s, 1960s, and 1970s. During the 1980s, unemployment started to increase again, reaching a peak of around 10 per cent in 1985. Using a dataset with information on 200 families over four consecutive generations, Mesters analysed the effect of the national unemployment rate on property and violent convictions.
Using generalized dynamic panel data models, and controlling for family-specific controls, macro-level trends and time-invariant family-specific effects, time-varying common effects and lagged offending outcomes, Mesters found that familial conviction rates are influenced by unemployment rates only in certain periods, and for certain crime types.

Specifically, only property convictions were influenced by the unemployment rate, and only during the post war period. We see no effect of the massive unemployment rate during the 1930s depression years. Based on previous research, researchers and policy makers generally assumed that economic downturns would lead to increased crime rates. This is however clearly not universally so: violent crime levels appear generally unaffected, and property crime rates only in certain periods.
The findings are of immediate relevance to current debates on crime rates, income equality, and unemployment. They also invite new theory development on the association between the economic climate and crime.
This study was carried out in collaboration with the Faculty of Economics and Business Administration of VU University, Amsterdam.
Mesters, G., Bijleveld, C.C.J.H. & Huschek, D. (2014). The effect of unemployment on crime in high risk families in the Netherlands between 1920 and 2005. In: Weerman, F. & Bijleveld, C.C.J.H. (Eds.). Criminal behaviour from school to the workplace. Untangling the complex relations between employment, education, and crime. London: Routledge, pp. 142-165.